Swithching Power

In [1]:
import pandas as pd
import mpld3
import numpy as np
import matplotlib.pyplot as plt
In [2]:
mpld3.enable_notebook()

df_waves = pd.read_csv('lossesHalfBridge.csv')
df_waves = df_waves.set_index('Time')
df_waves.head()
Out[2]:
I(L1) Iout I(IGBT1) count V2 V3 V6 count1 V13 V14 Vout vag Vsa1
Time
0.016667 -0.547685 0.749702 -0.547680 44 0.524595 0 1 0 0 0 0.749703 50.0 1
0.016667 -0.499183 0.748429 -0.499178 43 0.524595 0 1 0 0 0 0.748430 50.0 1
0.016667 -0.450679 0.747207 -0.450674 42 0.524595 0 1 0 0 0 0.747207 50.0 1
0.016668 -0.402172 0.746034 -0.402167 41 0.524595 0 1 0 0 0 0.746034 50.0 1
0.016668 -0.353663 0.744910 -0.353658 40 0.524595 0 1 0 0 0 0.744911 50.0 1
In [3]:
df_waves.tail()
Out[3]:
I(L1) Iout I(IGBT1) count V2 V3 V6 count1 V13 V14 Vout vag Vsa1
Time
0.033333 3.17670 2.09451 0.0 92 0.549097 0 0 0 0 0 2.09451 -0.000032 0
0.033333 3.12251 2.09556 0.0 93 0.549097 0 0 0 0 0 2.09556 -0.000031 0
0.033333 3.06832 2.09656 0.0 94 0.549097 0 0 0 0 0 2.09656 -0.000031 0
0.033333 3.01412 2.09751 0.0 95 0.549097 0 0 0 0 0 2.09751 -0.000030 0
0.033333 2.95993 2.09840 0.0 96 0.549097 0 0 0 0 0 2.09840 -0.000030 0
In [4]:
df_waves.columns

time   = np.array(df_waves.index)
pwm_s1 = np.array(df_waves['Vsa1'])
Iout   = np.array(df_waves['Iout'])
In [5]:
plt.subplot(211)
plt.plot(time, pwm_s1)
plt.subplot(212)
plt.plot(time, Iout)
plt.show()
In [6]:
from halfBridgeModule import detectCommutation, switchingLosesHalfBridge

Importando os modelos

In [7]:
import pickle

with open('models/SKM400GB125D.pklmod', 'rb') as filep:
    loaded = pickle.load(filep)
In [8]:
p_eon, p_eoff, p_erec, _ = loaded.values()
p_eDon = np.poly1d([0])

print('eon_params:', p_eon)
print('')
print('eoff_params:', p_eoff)
print('')
print('erec_params:', p_erec)
print('')
print('eDon_params:', p_eDon)

eon_params:           3            2
1.23e-10 x + 1.39e-08 x + 6.96e-05 x + 0.00464

eoff_params:           3            2
1.72e-10 x - 1.57e-07 x + 0.000105 x - 0.00104

erec_params:           3            2
7.25e-11 x - 1.32e-07 x + 8.53e-05 x + 0.00181

eDon_params:  
0
In [9]:
import numpy as np
import matplotlib.pyplot as plt

i = np.linspace(0, 1000, 200)

eon  = p_eon(i)
eoff = p_eoff(i)
erec = p_erec(i)
eDon = p_eDon(i)

plt.plot(i, eon*1e3, label='Poly On')
plt.plot(i, eoff*1e3, label='Poly Off')
plt.plot(i, erec*1e3, label='Poly Rec')
plt.plot(i, eDon*1e3, label='Poly D On')
plt.title('Polynomial Regression') 
plt.xlabel('$I_c$ [A]') 
plt.ylabel('$E_{on}$, $E_{off}$, $E_{rec}$ [mJ]')
plt.grid()
plt.legend()
# plt.yticks(np.arange(0, 110, 10))
# plt.xticks(np.arange(100, 650, 50))
# plt.xlim(50, 625)
plt.show()
In [10]:
P_sw = switchingLosesHalfBridge(pwm_s1, Iout, time[-1] - time[0], p_eon, p_eoff, p_erec, p_eDon)

# of (positive) Commutations per sec =  50462.62405645093
# of (negative) Commutations per sec =  50522.62717661318

Resultado Final das potências dissipadas de cada um dos semicondutores $S_1$, $S_2$, $D_1$ e $D_2$ , em Watts

In [11]:
mpld3.disable_notebook()

x = np.arange(4)
plt.bar(x, P_sw.values(), color =['red', 'green', 'yellow', 'blue'])
plt.xticks(x, P_sw.keys())
plt.ylabel('$P [w]$')
plt.xlabel('Semiconductors')
plt.grid()
plt.title('Switching Losses (Half-Bridge)')
plt.show()

mpld3.enable_notebook()
In [ ]: